Euler Movement Firefly Algorithm and Fuzzy Kernel Support Vector Machine Classifier for Keystroke Authentication
M. Rathi1, A. V. Senthil Kumar2

1Rathi. M, Assistant Professor in Department of Computer Technology, Dr. NGP Arts and Science College, Coimbatore since 2016.
2Dr. A. V. Senthil Kumar, Professor and Director of Department of Computer Applications, Hindusthan college of Arts and Science, Coimbatore.

Manuscript received on 27 August 2019. | Revised Manuscript received on 04 September 2019. | Manuscript published on 30 September 2019. | PP: 2267-2274 | Volume-8 Issue-11, September 2019. | Retrieval Number: K20630981119/2019©BEIESP | DOI: 10.35940/ijitee.K2063.0981119
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: User authentication can be successfully employed using keyboard typing patterns which is a form of behavioural biometrics. This modern method is highly analyzed for static authentication which refers to typing of fixed texts like ‘password’ and ‘pin numbers’. Most of the methods with respect to keystroke dynamics are restricted to the study of user’s activity involving fixed text. The formulated work concentrates on the investigation of the log of the user activity focused on the keyboard usage within the computer system through free text which refers to typing of texts throughout the login session. The Buffalo dataset is used in User Profiling Similarity Measurement (UPSM) stage and to recognize the time slice of the users, Euler Movement Firefly Algorithm (EMFA) is utilized. The typing behaviour is formulated in the form of time series in User Profiling Continuous Keystroke Authentication (UPCKA). Moreover the progression is made to user’s Continuous Authentication so as to predict unauthorized users with the consideration of the classifier called Novel Fuzzy Kernel Support Vector Machine (NFKSVM). The experimental results provide the enhanced performance by utilizing the formulated UPCKA in correlation with the NFKSVM classifier when compared with SVM and Iterative Keystroke Continuous Authentication (IKCA) techniques.
Keywords: Keystroke, Keystroke Time Series, Continuous Authentication, Buffalo dataset, User Profiling Similarity Measurement (UPSM) and User Profiling Continuous Keystroke Authentication (UPCKA).
Scope of the Article: